基于监测特殊机器人图像识别及射击准确度装置  

Image Recognition and Strike Accuracy Device Based on Monitoring Special Robot

作　　者：汪涌泉 李文嫄 马鲁霞 汪孝良 Wang Yongquan;Li Wenyuan;Ma Luxia;Wang Xiaoliang(College of Electrical Engineering&Automation,Shandong University of Science&Technology,Qingdao 266590,China;Heze Company of State Grid Shandong Electric Power Company,Heze 274000,China;Weihai Company of State Grid Shandong Electric Power Company,Weihai 264200,China)

机构地区：[1]山东科技大学电气与自动化工程学院,山东青岛266590 [2]国网山东省菏泽市电力公司,山东菏泽274000 [3]国网山东省威海市电力公司,山东威海264200

出　　处：《兵工自动化》2020年第7期89-92,96,共5页Ordnance Industry Automation

摘　　要：为解决机器人无法对连续不断改变的动态物体进行快速识别以及准确找到图像中心位置的问题,设计一种训练特殊机器人图像识别和监测射击准确度并进行实时报靶的装置。通过Mini PC输出和更换图像,讨论一种基于有限脉冲响应的多项式预测滤波器和改进的中值滤波相结合的混合滤波方法,对压力传感器的输出信号进行滤波降噪,STM32通过识别消噪和放大后的信号判断受击打的位置;并通过数码管进行显示。实验结果表明,该装置具有较高的灵敏度和准确性。To solve the problem that the robot can not quickly identify continuously changing dynamic objects and accurately find the center position of the image,a device is designed to train special robot image recognition,monitor shooting accuracy and report target in real time.First,output and replace the image with the Mini PC.Next,pressure sensor output signal polluted by additive noise are processed by mixed filter with polynomial prediction filter of finite impulse response and the improved median filter.Then,the STM32 determines the location of the hit by identifying the de-noised and amplified signals.Finally,the hit data is displayed by digital tube.The experimental results show that the device has high sensitivity and accuracy.

关 键 词：压力传感器 AD8230 STM32 多项式预测滤波器 中值滤波 

分 类 号：TP311.1[自动化与计算机技术—计算机软件与理论]